مصرف در مقابل مالیات بر درآمد زمانی که سرمایه گذاری سرمایه انسانی خصوصی کاملا قابل مشاهده است

This paper considers optimal taxation in an endogenous growth model where private education investments are imperfectly observable. Consumption taxation is better than labor income taxation for public provision of goods unless educational investment is completely unobservable. If subsidies are feasible for observed education investment, the consumption tax rate is independent of the degree of observability but the subsidy rate is higher the lower is the observability. If subsidies are not feasible, the consumption tax rate is lower the more limited is the observability. Optimal tax rates for goods that provide consumption and education investment simultaneously are below normal rates for observed pure consumption. Growth and welfare are positively related to (independent of) the degree of observability without (with) subsidies.

The conventional optimal taxation literature based on a representative agent focuses on efficiency arguments and draws conclusions in favor of consumption taxes or taxes on initial capital stock over income or labor income taxes; see, e.g., Atkinson and Stiglitz, 1981, Auerbach et al., 1983, Chamley, 1986, Cooley and Hansen, 1991, Devereux and Love, 1994, Jones et al., 1993, Judd, 1987, King and Rebelo, 1990, Lucas, 1990, Pecorino, 1993, Pecorino, 1994, Perroni, 1995, Rebelo, 1991, Summers, 1981 and Trostel, 1993. When the importance of human capital is recognized a key question arises: can governments distinguish private human capital investment from private consumption when a consumption tax is used? For many goods and services, the answer tends to be negative. A wide range of goods and services have both a pure consumption and human capital investment component. Major commodities like food, shelter, and clothing, for example, are all in this category — they are essential for bringing children up and also for maintaining the human capital of adults. Also, books, magazines, computers, radio, TV, and private lessons are recreational as well as educational. In fact, it is easier to make a list of things which have very little human capital investment aspect — e.g. tobacco — than it is to list all the goods and services which people use to build up or maintain their human capital. Thus, it is prohibitively costly to impose different taxes based strictly on actual uses, and it is realistic to take as a fact that governments have only a partial ability to distinguish private consumption from private educational investment. Owing to this fact, the publicly identifiable portion of private investment in education is generally exempted from a consumption tax or subsidized while the remaining portion is taxed. Clearly, taxing private educational investment to some extent under a consumption tax distorts individuals’ decisions. Thus, governments’ partial ability to differentiate private consumption from private educational investment poses some challenging questions. Firstly, can a consumption tax do better than other taxes in terms of welfare when private educational investment is (at least partly) subject to this tax? Secondly, what are the implications of the degree of this partial ability for growth and for the level of government expenditure which should, and perhaps will, be chosen? Much work on optimal taxes focuses on lump-sum transfers of tax revenue to individuals. In practice, however, there are substantial uses of tax revenue to provide goods and services, including public consumption and investment in education in many countries. Most of the OECD countries spend more government revenue on the provision of goods and services than transfers net of social security.1 Using the neoclassical growth model, Krusell et al. (1996) have recently shown that if government outlays are used for redistribution through lump-sum transfers, then income taxes are not necessarily worse in welfare terms, and may even be better than consumption taxes. In the political equilibrium of their model, income taxes are attractive precisely because they are more distortionary, since this implies that low equilibrium transfer levels will be chosen. In the case of public in-kind provision, however, they show that the best tax is the least distortionary one, which in their model as in many others is the consumption tax. In recent endogenous growth models that include human capital and the labor-leisure choice, consumption taxes are found to be better than taxes on physical capital income or labor income (e.g., Lucas, 1990, Pecorino, 1993 and Perroni, 1995).2 It is interesting to ask what happens then if the consumption tax is distortionary owing to governments’ partial ability to distinguish between private education investment and private consumption. Our purpose in this paper is to consider optimal flat-rate taxation for public provision of goods and services when there is partial public knowledge about the uses of private goods for consumption vs. investment in education. The limited observability of private spending appears in different forms with different types of expenditures (pure consumption, pure investments, or mixed), and we will investigate these different cases. In doing so, we assume that public and private educational investments are not perfect substitutes. We also consider endogenous growth of per capita income through human capital accumulation to capture the dynamic growth effects of taxes and public expenditure as in some recent optimal taxation literature (e.g., Barro, 1990, Lucas, 1990, Pecorino, 1993, Perroni, 1995 and Stokey and Rebelo, 1995). Our attention is confined to taxes on consumption vs. labor income since our model abstracts from non-human capital. In our model, both income and consumption taxes exert opposing forces on welfare and growth. On the one hand, both cause an intertemporal distortion due to their handling of the direct costs of human capital investment. (Indirect costs, that is forgone earnings, are implicitly deductible from either tax and therefore do not cause a problem.3) Under a pure labor income tax there is no deduction for direct costs, and with partial observability some direct costs will be taxed under a consumption tax. These elements raise the cost of investment and reduce after-tax rates of return, tending to reduce welfare and the growth rate. On the other hand, using the tax revenue in part to provide public investment in education tends to raise welfare and the growth rate. The net effects on growth and welfare are of central importance. In particular, the degree of governments’ partial ability to identify the actual uses of a private good will be critical to size the distortion of the consumption tax, and hence may affect the optimal rate of the consumption tax, and the optimal mix of income taxes and consumption taxes. The main results in this paper are as follows. The optimal tax solution depends on the government’s ability to distinguish private human capital investment from private consumption. If it has some ability in this regard, only a consumption tax should be used to finance public provision of goods and services. If subsidies are feasible for observed education investment, the consumption tax rate is independent of the degree of observability but the subsidy rate is higher the lower is the observability. If subsidies are not feasible, the consumption tax rate is lower the more limited is the observability. Optimal tax rates for goods that provide consumption and education investment simultaneously are below normal rates for observed pure consumption. Growth and welfare are positively related to (independent of) the degree of observability without (with) subsidies. The growth effect of a consumption tax for public goods is also discussed. The remainder of this paper is organized as follows. The next section introduces the model where private investments in education are fully observed. Section 3 investigates the model with limited observability of education investment. Section 4 makes some extensions. The last section gives some concluding remarks.

This paper has investigated optimal taxation in an endogenous growth model where governments can only distinguish partially between private human capital investment and private consumption. We have shown that optimal taxation for public provision of goods and services depends on the government’s ability to distinguish private human capital investment from private consumption. The government should use only the consumption tax instrument as long as some fraction of private investment in education can be identified. The latter should ideally be subsidized. If subsidies are not feasible, the lower is the government’s identifying ability, the lower is the optimal consumption tax rate. With subsidies, the optimal consumption tax rate is independent of the degree of observability but the subsidy rate is higher the lower is the government’s observability. If the government is totally unable to identify the uses of private goods, then it can use either the consumption tax, or the labor income tax, or any mix of them. Optimal tax rates for goods with mixed benefits of consumption and human capital investment, and for fully indistinguishable pure consumption and investment are positive but below normal rates for observed pure consumption. It can also be shown that without subsidies, the growth rate and welfare are positively related to the government’s identifying ability, through numerical solutions with general values for this ability as well as analytical solutions with special values for this ability. With subsidies, the growth rate and welfare are independent of the degree of the observability if there is no additional cost associated with the provision of subsidies. These results may have some policy implications. Due to governments’ partial knowledge about actual uses of private goods, the advantages of consumption taxes over labor income taxes are smaller than those suggested in the literature. In practice, the provision of subsidies to observed educational investment involves some additional administrative costs. If these costs are small, subsidizing observed educational investment may be the best option. When such costs are substantial, exemption of observed educational investment from tax may do better although its gain tends to be smaller the more limited is the observability. In either case with or without subsidies, replacing income taxes with consumption taxes will yield smaller gains than the proposed ones in previous work. Finally, in the real world there may be good reasons for the exemptions or lower consumption tax rates often observed on goods and services which may contribute to human capital investment as well as consumption. The gains of switching from income to consumption taxation may fall further in a model that includes physical capital. As pointed out by a referee, replacing income taxes with consumption taxes will increase the tax rates required to raise a given amount of tax revenue because the tax base becomes smaller, especially when physical capital income taxes are considered as well. If the government is sufficiently poor at identifying investment in human capital and if administrative costs of subsidies are too great, net taxation of inputs into human capital may therefore rise when income taxation is replaced by consumption taxation in models with observed investment in physical capital. This suggests the possibility that a move from a comprehensive income tax to consumption taxes may generate very small welfare gains or even welfare losses. Investigation of this issue would be a good area for future research. In a model like ours, without non-human capital, the traditional contrast between an income tax which distorts saving, and a consumption tax which is neutral in that regard, does not apply. This naturally tends to produce more favorable growth and welfare effects under an income tax than when non-human capital is included. But while we abstract from this traditional contrast, we also bring in that between the distortionary impact of an income tax on investment in human capital vs. the less distortionary impact of a consumption tax. Recent literature has stressed the general importance of human capital and its specific role in the growth process. Our results should therefore be of interest even though they abstract from the form of income tax distortion which was formerly considered to be of central interest.